Widening the diversity of genetic databases

The diseases that we succumb to are frequently the product of the interaction between the environment we inhabit and the genes that drive our biochemistry. So two people can live side by side, in the same place, eating the same food and doing the same jobs yet have totally different disease risks. But before we can provide tailor made interventions, specific to a person’s make-up, we need to know what the disease associations are. And we’re beginning to get there. Biobanks are collecting epidemiological, clinical and genetic data allowing us to link genes to disease and environmental risks. But many of these archives are centred on quite narrow slices of the human population, which can limit their utility to certain populations, and means that many genetic variants, and their attendant protective or deleterious influences, are never uncovered. As she explains to Chris Smith, Bana Alamad, at the University of Oxford, has been looking at the role that diversity plays in our aspiring vision for personalised medicine…

Bana - We hear the term personalised medicine a lot and we know that understanding genetic information can really help us realise this dream of medicine becoming personalised and tailored specifically for the benefit of individual patients. And when we started thinking about this review, we really wanted tolook within the immune related disease space and really propose what we think the challenges we're going to face and we are facing are, but also suggest the essential direction that the field of genomics needs to head in, in order to actually realise this dream of personalised medicine.

Chris - Now, if you and I had been chatting maybe a decade ago, the conversation would probably have dwelled really heavily on the cost. But now the cost of doing genetic analysis has gone straight through the floor, hasn't it? It's a fraction of what we were spending when we did, for instance, the Human Genome Project. So what does the problem or the kind of obstruction, the obstacle, what does it now look like for doing projects like this? Where are the bottlenecks?

Bana - Mm, completely true. So as you mentioned, high throughput genetic sequencing is so cheap now in comparison to where it's been. It's really revolutionising our ability to identify genetic variation between population in the context of disease. And this is why our field is really very fast paced, I would say. But I think there are key challenges and key areas that we really need to think about. And the first thing is to really think about the interplay between genetic information and environmental factors and how this interplay affects determining the risk of susceptibility to immune disease. The second really important thing is considering genetic variation across diverse populations. Not just including everyone for the sake of diversity, but because it's super important to help us understand disease aetiology and what different variants can mean and how they can lead to diseases and so on. So I'd say these are my top two.

Chris - So if we just focus on one very homogeneous population, which you tend to get in say a western country, there are very likely going to be populations in many other geographies facing the exigencies of the environment there, lifestyle there, which may well mean there are some interesting genetic forms that would otherwise escape and may give us insights into all new kinds of mechanisms of disease or even disease treatment. Is that what you're saying? By that diversity point?

Bana - Yes, but I'm also saying that currently we have limited clinical utility when it comes to genetic-based diagnosis and risk susceptibility and so on in these non-Caucasian populations. Really studying all of this variation in these specific populations can really help us highlight and understand mechanisms and the way disease progresses that we won't be able to do that if we keep only studying homogeneous population.

Chris - What's your solution to achieving that then? What do you think we should do? Should big charities or research fund sources, or even governments of many of these countries just put up huge amounts of money then and say, right, we are going to start genetically sequencing a good cross section of our populations so that we end up with a full database?

Bana - Well, that's one option. So I think large scale biobanks that are dedicated to populations, to different various diverse populations, is crucial because you need to have an understanding of as much variation as you could capture, which means that we can identify variation that is common and also extremely, extremely rare. But I also think, beyond that, if we really want to think about what we want and where our field should head, we need to think about capacity building because I can't just go and sequence, say, X population because they are different and they are understudied without thinking about capacity building in designing and implementing research that is up to standard with these international methods that we're using. So it's beyond the data in my opinion. It's more about capacity building, setting up infrastructure and also representation.

Chris - Some people would turn around and they'll say, 'well, resources are limited. It all comes down to money. The world could always use more of it, and there's only therefore so many things we can spend it on.' Is this the best thing we can spend the resources on right now or should we focus on putting fires out and worry about tarting up the buildings later? Because this is very much a capacity building thing, it is not an instant fix.

Bana - Hmm. I think this is a very good point. In general, medicine has always been very reactive. Someone gets sick and then it takes us time to diagnose that it takes us time and resources to go through understanding the prognosis. And it takes us time to figure out the proper treatment. And this is very expensive. So if we want to use the money argument, I would say that it's extremely expensive to treat someone who has advanced stage of disease, especially when you think about many, many people and you think about the very wide large scale. So I think through investing in proper infrastructure and then key targeted genetic approaches, we can really start improving this case and really shifting towards this concept of personalised medicine. And I think, in the long run, this is going to save us money...